The validity of actigraphy to make this distinction relative to polysomnography depends on several factors, including the population being studied and the time-of-day. In healthy, normal young, and middle-age adults sleeping at night, the agreement between actigrahpic and polysomnographic assessment of sleep time is generally 90% and better. 1 In contrast, in the elderly and patients with sleep disorders, the percent agreement generally is less, sometimes 80% or lower. 1 The discrepancy is in both directions with over-and under-estimations of sleep time. Some elderly and patients often lie motionless while awake producing over-estimation of sleep time, while other patients may have many body movements or leg movements during sleep producing under-estimations of sleep time.As to time-of-day, detecting sleep episodes with actigraphy in the context of normal waking activity has not been successful. In healthy young adults actigraphy differentiated sleep and wake at night, but it did not do so during daytime napping. 2 The agreement between actigraphic and polysomnographic identification of sleep was 40%. Periods of inactivity while reading or watching TV could not be differentiated from daytime nap episodes.While actigraphy may not detect daytime sleep episodes, it may differentiate increasing levels of daytime sleepiness as reflected in reduced overall activity or increased brief periods of inactivity. Reducing sleep time in healthy normals increases the level of daytime sleepiness in a linear manner as measured by the Multiple Sleep Latency Test (MSLT). 3,4 Increased daytime sleepiness then may be associated with reduced activity. We know of no previous attempts to document with actigraphy differing levels of daytime sleepiness due to reduced bedtime the previous night. Since previous studies have shown increased sleepiness and reduced performance following sedating drugs, the antihistamine diphenhydramine was used as a positive control in this study. 5,6 Thus, this study compared the effect of sleep loss, due to reduced sleep times, with the effect of diphenhydramine 50 mg on the MSLT and waking actigraphy. It was hypothesized that sleep loss the previous night would reduce activity levels during the following day. Each day the volunteers wore actigraphs from 0700-1800 hrs. Decreasing TIB was associated with decreased daily mean sleep latency on the MSLT with 4 and 0 hrs differing from 8 hrs and each other. Daytime activity also was reduced by the reduced prior TIB. Increased inactivity relative to the 8 hr TIB developed between the 4 hr and 0 hr TIBs, with 4 hrs differing from 0 hrs, but not 8 hrs.
Accepted for publication June 2000Diphenhydramine 50 mg reduced mean daily sleep latency and increased percent inactive time relative to placebo. On the MSLT diphenhydramine was intermediate to 4 hr and 0 hr TIB and on actigraphy it was similar to 0 hr TIB.
Conclusions:The difference in the effect of diphenhydramine on these actigraphy and MSLT may reflect the different sensitivities of the measures.